onsite
Perception Engineer
Perception Engineer
The Perception Engineer will develop and deploy CV/ML algorithms, deep learning models, and vision-language models for real-time understanding of ego vehicle surroundings. This role involves creating and optimizing data pipelines, conducting research in perception algorithms, and utilizing multi-modal sensor fusion for autonomous driving applications.
About the role
Team Introduction
The Perception team is responsible for developing CV/ML algorithms to understand the surroundings of ego vehicles in real time, including obstacle detection, online mapping, and scene understanding. The perception system takes inputs from multiple sensors across time, and outputs various obstacle-level and scene-level understandings for downstream modules such as prediction and planning, etc.
In the Perception team, you will:
- Develop and deploy cutting-edge perception and deep learning models, including computer vision models, vision-language and large language models (VLMs and LLMs), for real-time integration into our system.
- Create and optimize data pipelines for data mining, labeling, training, and evaluation, supporting comprehensive onboard development.
- Conduct advanced research in perception algorithms to address complex, real-world scenarios, focusing on areas like object detection, segmentation, road understanding, and motion prediction.
- Leverage multi-modal sensor fusion techniques to build robust architectures tailored for autonomous driving.
Skills
CV/ML algorithmsobstacle detectiononline mappingscene understandingdeep learning modelscomputer vision modelsvision language models (VLMs)large language models (LLMs)Data PipelinesData Mininglabelingtrainingevaluationperception algorithmsobject detectionsegmentationroad understandingmotion predictionmulti modal sensor fusion